Optimal dismantling of criminal networks. A perspective from the mathematical and computational modeling

This work deals with the study and comparison of different strategies for the optimal dismantling of delinquent networks, which aim to optimally identify the most relevant individuals in the network. The strategy of greater complexity that we have studied here, is based on the Katz-Bonacich centrality criteria as a measure of influence of the individuals in the network. This results in an NP-hard type of problem, therefore, in order to apply that criteria, we must use heuristic methods which allow us to find approximate solutions. In particular, the methods used in this work are the Monte Carlo and greedy algorithms. We compared their performance against less sophisticated strategies and we were able to find that these algorithms perform relatively better, which contributes to improve our understanding of these approaches. In addition, we discuss a model that was recently introduced, which justifies the use of Katz-Bonacich centrality from the point of view of game theory on networks.

• Autores: Sarmiento Bahoque, Tomas Angel; Cantillo Palacio, John Fredys; Realpe Gómez, John Eduardo; Montoya Martínez, Javier Antonio
• Fecha de publicación: 2016
• Institución: Universidad EAFIT
• Repositorio: Repositorio EAFIT
• Idioma: spa
• OAI Identifier: oai:repository.eafit.edu.co:10784/11293
• Acceso en línea: http://publicaciones.eafit.edu.co/index.php/ingciencia/article/view/3293; http://hdl.handle.net/10784/11293
• Palabra clave: Complex systems; computational modeling; network models; game theory; delinquent networks; statistical mechanics; ICT; Sistemas complejos; modelado computacional; modelos en redes; teoría de juegos; redes delincuenciales; mecánica estadística; TICs
• License: Copyright (c) 2016 Ingeniería y Ciencia | ing.cienc.